Abstract: Renewable energy is widely distributed in remote or coastal areas. In order to improve the economic efficiency and consumption rate of such renewable distributed generators (DGs), this paper presents a robust planning model for AC-DC distribution system, where DGs are connected as a DC network and provide power supply for AC system through collection lines. The planning model is decomposed into an investment master problem and an operation sub-problem. The optimal planning decisions determined in the master problem are delivered to the sub-problem, where the operating cost under the worst N-1 contingency are calculated based on Dual technology and Big-M method. The second order conic programming (SOCP) based column-and-constraint generation (C&CG) algorithm is applied to solve the two-stage robust planning model. Numerical results on a revised AC-DC test distribution system exhibit the effectiveness of the proposed model.